Field of the Invention
The present invention relates to image capturing apparatus and control method thereof.
Description of the Related Art
Many image capturing apparatuses such as video cameras employ methods in which a video signal created by an image sensor is converted into digital data, the data is subjected to various types of signal processes and then encoded using a lossy compression technique, and the encoded data is recorded into a recording medium. However, some professional cameras, such as those used to film motion pictures, employ methods in which the video signal created by the image sensor is recorded after undergoing lossless compression or is recorded as image data in the RAW format, which is an uncompressed format (this will be called “RAW image data” hereinafter). This method enables a developing process in which a user can specify a variety of parameters to be carried out after the recording, so that the user can create a video aligned with his or her intentions.
When recording RAW image data, it is typical to use a technique in which developing parameters are also recorded as added information (metadata) and can then be used in the developing process. For example, a technique is known in which when performing a magnification process realized through digital zooming, a magnification rate is saved as metadata and the RAW image data is recorded as-is (Japanese Patent Laid-Open No. 2009-147743). Meanwhile, there is an increasing need to handle high-resolution video data such as 4K and ultra-high-definition formats, which tends to dramatically increase the size of the recorded RAW image data. In response to this, a technique is known in which data amounts are reduced by resizing images to sizes that are sufficient for the majority of users (Japanese Patent Laid-Open No. 2007-166551).
Incidentally, when recording RAW image data, the minimum required image correction is carried out, such as noise removal using pixels in an optical black region (“OB region” hereinafter) output from the image sensor, a clamping process (setting a black level), or correction for missing pixels. However, with the increase in processing amounts required by high-resolution video signals such as 4K and ultra-high-definition, it is desirable to suppress the amount of processing performed on the video signal and reduce the scale of the circuitry and so on in the image capturing apparatus by employing a configuration that carries out the minimum required image correction in the post-recording developing process as well.
It is necessary to record RAW image data including the OB region in order to realize a configuration in which the minimum required image correction is carried out in the post-recording developing process. However, how to handle the OB region data becomes an issue when carrying out resizing processes such as digital zooming, switching aspect ratios, dynamic extraction position changes caused by digital image stabilization, and so on. In particular, if the OB region is changed dynamically in response to dynamic switches in the angle of view due to digital zooming, dynamic changes in extraction positions due to digital image stabilization, and so on, major changes in image correction can be expected to occur from frame to frame, depending on the quality of the image sensor. Although the aforementioned Japanese Patent Laid-Open No. 2009-147743 and Japanese Patent Laid-Open No. 2007-disclose techniques for carrying out magnification processes on RAW image data, those documents do not disclose techniques for applying magnification processing to RAW image data that includes an OB region.